The train is practically the same as that used in the tests of the 4-6-2 type locomotive, and had an average weight of 288 tons, including passengers and lading on the superheater tests and 289 tons on the saturated steam tests. The former locomotive weighed 189 tons as compared with 186 tons for the latter. The distance was 96.82 miles and the schedule allowed two hours and forty-eight minutes. The weather in each case was fair, wind light and rail good. On the first trip with the superheater locomotive there were ten slow orders in addition to nine stops and 4 min. 50 sec. were lost on the trip. On the second trip there were but five slow orders in addition to the nine stops and the run was made in three minutes less than schedule time. On the third trip, with the saturated steam engine, there were five slow orders and nine stops, and fifteen minutes were lost. On the fourth trip there were five slow orders and nine stops and three and a quarter minutes were lost. The coal in each case was an excellent grade of run of mine. The following Mikado Locomotive Tested Both With and Without a Stoker on the Buffalo, Rochester & Pittsburgh. Atlantic Type Passenger Locomotive with Brick Arch and Superheater on the Buffalo, Rochester & Pittsburgh. Heating surface, superheater Superheater, number units.. Grate area ..... Saturated. 184,000 lbs. 164,000 lbs. 21 in. x 28 in. 57 in. 200 lbs. 36,827 lbs. 70 in. 190 sq. ft. 2,672 sq. ft. 2,862 sq. ft. 54.4 sq. ft. Superheated. 194,000 lbs. 173,500 lbs. 21 in. x 28 in. 57 in. 200 lbs. 36,827 lbs. 70 in. 218.8 sq. ft. 2,154.6 sq. ft. 2,373.4 sq. 460 sq. ft. 28 ft. 54.4 sq. ft. The same excellent quality of run of mine coal was used in each of the four runs all of which were hand fired. The weather and rail conditions were good and the same for each run. In each case a helper was used from Falls Creek to McMinns Summit, a distance of four miles. On the runs with the superheater locomotive this consisted of a consolidation locomotive with 21 in. x 28 in. cylinders, 57 in. drivers, and a tractive effort of 35,827 lbs. A larger helper was used on saturated steam tests, this consisted of a 22 in. x 28 in. consolidation with 56 in. drivers and 200 lbs. steam pressure. Its tractive effort is 41,140 lbs.. A pusher engine was attached at Clarion Junction and helped the train to Freeman, a distance of about 171⁄4 miles. This was a decapod with cylinders 24 in. x 28 in., 52 in. drivers, 200 lbs. steam pressure, and a tractive effort of 52,730 lbs. The same locomotive was used at the same places on each of the runs and no deduction is made for the work that either the helper or pusher did in handling the train, nor are their weights or coal consumption included in the results. The following table shows the train handled on each of the runs : The result of these tests very clearly shows the advantage of the superheater and brick arch for both passenger and freight service and the Buffalo, Rochester & Pittsburgh is now engaged in equipping other locomotives with this apparatus as rapidly as possible. INTERNATIONAL ASSOCIATION FOR TESTING MATERIALS. The International Association for Testing Materials held its sixth congress in the Engineering Societies building in New York, September 3 to 7, as mentioned in the Railway Age Gazette of September 6. PAPERS ON RAILS. Among the papers presented, a group of seven on rails are of direct interest to railway men. The first was on "Welding the Blow-Holes and Cavities in Steel Ingots" by J. C. Stead. He attributed blow-holes to the presence of gas, which is either initially present or is formed by chemical reaction brought about by a fall of temperature. He outlined tests which showed that an ingot known to contain blow-holes, on reheating to 2,000 deg. F. proved to be as sound as ingots which had not contained blow-holes, and he concluded that under the ordinary treatment to which honeycombed ingots of steel are subjected in heating and rolling, internal cavities or blow-holes are perfectly welded providing there is an absence of sulphide segregations. The paper by A. Mesnager on "Means of Foreseeing Ruptures of Rails" described a method of inspection practically amounting to polishing and examining microscopically; a process which was criticised as being impossible of practical application. J. P. Snow read a paper on "Some Features of the American Steel Rail Situation," in which he commented on the development in rail sections, and of the tendency towards the more general use of open hearth metal. In his paper on "Some Features Associated with the Test of Steel Rails," James E. Howard repeated a suggestion which he has previously made regarding the cold rolling of rail heads. He also suggested as a means of detecting slag laminations, a test in which the base of the rail would be bent in a crosswise direction. P. H. Dudley, in his paper on "Testing Rails for Elongation and Ductility of the Metal Under the Drop Testing Machine," gave a brief resume of a series of tests made under mill conditions for measuring the elongation of the metal. It was stated that the adoption of these tests had resulted in a modification of mill practice, which will materially increase the elongation obtainable and that it has proved possible to hold the elongation within the limits of 18 to 20 per cent. M. H. Wickhorst, in a paper on "American Research Work on Rails Conducted Jointly by Railways and Steel Manufacturers," reviewed railway and mill conditions and the work that is being done to secure better rails. The paper by Robert W. Hunt, on "Insuring Soundness in Steel Rails," is published elsewhere in this issue. HARDNESS TESTING AND WEAR. That feature which was brought most prominently to the front in section A was the structural peculiarities of metals and the effect of these peculiarities on the service to which the metal was to be devoted. This keynote was struck in the first paper presented on "Hardness Testing and Resistance to Mechanical Wear," by E. H. Saniter, which showed that while a high Brinell number for hardness may be expected to give better wear than a lower one, there are so many exceptions that its use for the purpose of indicating wearing properties is unreliable as far as the methods of wear testing reviewed are concerned. The relation of either Brinell tests or wear tests to wear in actual practice requires investigation. The greater reliability of the wear test as an indicator of wear is emphasized by the test on Hadfield's manganese steel which gives, with a low Brinell number, the best wear number of all the steels tested. TESTS OF WEAR OF STEELS. Felix Robin, in a paper on this subject, showed that in the case of annealed carbon steels the wear is not proportional to the contents in pearlite and that increased fineness of the particles, cold work and the presence of phosphorus increase the resistance to abrasion; the presence of silicon and manganese diminish it frequently. In cast metal the resistance grows with the phosphorus contents and with the percentage of iron carbide. The hardened steels can hardly be distinguished by this process; on the contrary, tempered steels lend themselves to this examination without any difficulty. The best resistance in hardened steels seems to be characteristic of the finest martensites. Chromium increases the resistance of annealed steels noticeably; it has little effect on hardened steels. HARDNESS OF STEEL. Captain C. Grard, in a paper on "Research on the Hardness of Steel," proposed to substitute a hardness test with the Brinell ball for the ordinary tensile tests. He maintained that if we multiply the Brinell number by a coefficient that can be obtained the result will be the tensile strength of the metal and that this coefficient undergoes but little variation in the different kinds of carbon steel. This fact can experimentally be verified in the ordinary way. It is of interest because it admits of simplifying tests. It will suffice to measure the diameter of the ball impression, to deduce the hardness figure from this impression, and finally to multiply this hardness number by the coefficient of proportionality in order to arrive at the tensile strength. These operations are simple and they dispense with the necessity for the always laborious preparation of test specimens. But the coefficient of proportionality is not absolutely constant. It undergoes slight variations with different grades of steel, variations which will be interesting to enquire into. TESTING STEEL TUBES. A new method of testing steel tubes was suggested by C. Fremont, who said that in welded tubes a certain amount of overlap at the joint must be insisted on. In connection with tests on welds, he has shown that, even in the best welds, there is only a kind of adhesion, of very feeble strength, which explains the numerous cases of fracture of steam pipes through the weld giving way. He had found that the overlap in certain steam pipes was equal to fifteen times the thickness of the metal, whilst in other cases the two edges were simply held together by a butt-joint weld without any overlap at all. In all pipes whose accidental breakage may form a source of danger, no information on the strength of the metal is afforded by static tests (tensile tests, distension, bending, crushing, etc.). For instance, a steel pipe, which successfully passed various static tests, proved extremely brittle under the impact test on nicked test pieces, the latter breaking under a moderate force and without any deformation of the metal. Viewed in the microscope, under a low power, 50 diameters, for instance, a polished section of the tube was found to be contaminated throughout with impurities which were revealed more clearly when etched with iodine. It is well known that such impurities may contribute toward rendering the metal brittle; and moreover they are also the cause of pustular corrosion, especially when the metal is in contact with a material that possesses more or less oxidizing properties, such as feed water. steel containing these impurities should not be used for making the various parts of boilers, collecting pipes, tubes, shells, etc. Hence If these impurities are distributed in the metal in such a manner that they accumulate very close to the surface, oxidation penetrates into the deeper regions, and the metal is then pitted in the form of galleries. For testing a steel tube, Mr. Fremont cut off a ring 8 mm. broad, from each end. Each of these rings is examined under the microscope for the presence of piping or of impurities large enough to produce pustular corrosion; and also, in the case of welded tubes, to verify the apparent quality and amount of overlap of the weld. Then, each ring found to be good under this preliminary examination is flattened and doubled into U-section test pieces-this shape being given them to ensure sufficient stability under the impact test-and are cut off, either with the saw or with the shears, this cold working at the ends having no effect on the portion tested. The doubled test piece, about 35 mm. in length, is nicked on both limbs of the U by a saw cut, 1 mm. long and 1 mm. deep, and subjected to the impact test. The force required to break the test piece should not be less, for instance, than 745 foot pounds per square inch of the initial sectional area at the break. WELDING STEEL. Max Bermann read a paper on "The Nature of Welding Different Kinds of Steel and its Practical Application," from which he drew the following conclusion: Summary and Practical Rules.-The perfect union of pieces of steel by welding is possible in the solid condition, but it is rarely realizable and can only be attained with certainty when the "Komm" method* is applied. The crust of oxides which prevents the complete union of the smallest particles on the surfaces is rendered metallically pure and freed of its oxygen by the reducing action of the steel, and the cohesional union is thus facilitated. The reduction of this crust of oxide proceeds all the more rapidly, the higher the temperature of welding and the greater the relative proportion of the reducing elements present in the steel. The welding temperature is the highest temperature at which the weldable steel remains malleable. The pressure which is required to effect cohesional union of the surfaces is relatively small at the welding temperature, but it must be sufficient to effect a very slight deformation at the seam and should be applied in accordance. The weld will be perfect when the parts united will not separate along the seam under alternating torsion or bending though continued until fracture ensues, but stronger pressure should be enacted and should be of the force when they break across like solid whole masses. If, however, the seams separate under stress as indicated, and if these surfaces are free from slags and cinders, the weld must be regarded as practically successful, because such a weld will successfully bear tension or bending if in one direction only. Results of Importance for Practice.-The weldability of the steel is determined by heating it up to sparking white glow and by hammering it immediately afterwards. When the steel remains malleable at this heat, it will also be good for welding. The heating of the pieces which are to be welded should take place in an oxidizing flame; otherwise the steel will take up carbon from the coal in the hearth and be converted into a nonwelding modification. The areas of the surfaces in contact should be made as large as possible. It is also recommended to shape the ends like wedges fitting into one another, and to take care to effect intimate contact of all the points on both sides of the wedges. When an intimate contact between the two surfaces has been provided before heating up to welding temperature, even a rela *The union of the two pieces is produced within the fire, where the two conically pointed ends are in contact. The crust of cinders is removed by blows applied to the butts which project from the hearth, and since the two surfaces which are thus kept pure cling to one another at once and unite, there is no opportunity for a renewed oxidation. Rods which are welded in this way do not separate under alternating torsional bending along the faces welded, but they break across the axis of the rod, just like rods which have not been welded. M tively large crust of cinders will not prevent the realization of a good weld, provided the steel be itself readily weldable. The pressure which is required to bring the two surfaces into good contact should be directed towards the center of the faces. It should only penetrate up to the middle and then be directed towards the marginal zone in order to facilitate the ejection of the slag. When. the surfaces have begun to stick together, a suitable pressure should be applied. which is used in forging pieces of corresponding dimensions. The test of the soundness of a weld by alternating torsion or bending at ordinary temperature should mainly be applied when it is a question of passing weldable steels in acceptance tests. The reagents to be used for the welding of steels which are not themselves weldable consist particularly of materials which form slags and of iron turnings, also of very thin wires; they help to bring about a union between the faces and to dissolve the oxides. These particles of iron should, however, be applied in a very finely distributed state; otherwise they will be more injurious than useful, because they will interfere with the squeezing out of the slag. Summary. The author arrives at the conclusion that the presence of manganese is favorable for the success of the welding, because the manganese reduces the crust of oxides. RAILWAY CAPITALIZATION AND TRAFFIC. BY H. T. NEWCOMB, Of the Bar of the District of Columbia. chine, that is to say, every result of indirectly productive labor is capital and its efficiency is measured by the amount of direct labor with which it dispenses. In other words, the efficiency of capital is in inverse ratio to the quantity of labor required to produce a given volume of output by its operation. Transportation is a productive service of the most fundamental character, and an essential part of every productive process. It is a commodity, and like other commodities may be produced by the use of capital or, crudely and in very limited quantities, by labor alone. And the processes of producing transportation, like those of producing other commodities, have become increasingly efficient as more and more capital has been devoted to them. American railways carry a ton of freight one mile, on the average, for about three-fourths of one cent, and this sum, so small that the ton must be carried two hundred miles to pay $1.50 for the day's work of the most unskilled laborer, must suffice to pay all the labor, much of it most highly skilled, that is necessary, and to afford a fair return upon all the capital employed. Also, the American railways, in the fiscal year 1910, the latest for which official data are available, carried 255,016,910,451 tons of freight and 32,338,496,329 passengers one mile. These enormous aggregates must be reduced to averages to be comprehended at all. The averages per capita of population and per railway employee are: Economic science recognizes two general classes of commodities, which are, first, those like food and clothing, directly capable of satisfying human wants, and, second, those, like flour mills and sewing machines, useful in the production of commodities of the first sort. Commodities of the second class, being unsuited for direct consumption, or segregated from the mass of goods intended for direct consumption, and devoted to the production of other goods become, by reason of the use to which they are put, capital. Labor may be devoted to the production of the one class of commodities or the other; if to the first class it is directly productive, if to the second class it is indirectly productive. Obviously the only sound reason for the indirectly productive employment of labor that could exist would be that in consequence of such employment the sum of commodities capable of directly satisfying human wants finally produced by a given expenditure of effort is augmented. Indirect production, as it involves the use of capital, that is of tools, machinery, etc., is capitalistic production. Under modern methods of industrial association all production has become capitalistic in greater or less degree; that is to say, there is left no vestige of methods so crude that the final directly productive effort is not lightened by the utilization of the results of some indirectly productive labor that has gone before. In other words, even the poorest laborer uses some tool or tools; there are no remaining processes of producing articles to satisfy human needs that do not make some use of capital. The industrial marvels of the nineteenth and twentieth centuries, those triumphs of the arts of peace which have enabled Europe and America to support vastly multiplied masses of population and give to each individual of those masses opportunities for comfort far beyond those enjoyed by the most fortunate of their predecessors, are the specific results of the progressively increasing use of the capitalistic method of production. The productivity of human energy as expressed by the production of goods for direct consumption, has expanded in a ratio far beyond the possibility of admeasurement, because an increasing share of human labor has been devoted, not to the direct, but to the indirect, production of such goods. For example, labor of one man, devoted to the production of textile machinery, in its ultimate result, produces more clothing than the labor of a thousand men expended for the same final result, but compelled to work without machinery. Every tool or ma A consciousness which comprehends the fact that 2,773 tons of freight were moved one mile by American railways in 1910 for each man, woman and child of the country's population can, perhaps, contrast with that result, accomplished by this capitalistic process, the capacity of the strongest man to move the heaviest load of which he is capable, and from this contrast, vaguely imagine the state of industry and civilization that would remain if this method of capitalistic production suddenly became unavailable. Or, in imagination, the results of the existing method may be contrasted with the cruder capitalistic processes of transportation which the railways have supplemented and superseded -for the wheelbarrow, the horse and wagon and the highway are each capital. The railway itself has not failed to progress in efficiency-its efficiency when introduced was high only with relation to the means of transportation which preceded it; its growth in efficiency has been by the progressive measure in which it has enabled capital to take over a steadily increasing share of the tasks formerly devolved upon labor alone. From 1880 to the present time this progress can be read in the official statistics. Note the following. No account is taken in the foregoing of the additional services performed in the transportation of persons, which have also increased greatly both in their total volume and in the average per employee. For present purposes it is, however, unnecessary to complicate the calculations by extending them to include passenger services which may not improperly be regarded as a bye-product incidental to the principal services which are those rendered in the transportation of property. If the higher productivity of labor, secured by throwing an increasing share of the work upon capital, had not changed the ratio of tonmileage to number of employees, as compared with 1880, the number of employees required by the traffic movement of 1910 would have been 3,302,772 instead of 1,699,420; if such productivity had been no greater in 1910 than in 1890 the number of employees in the later year would have been 2,507,442; if no greater than 1900 it would have been 1,832,808. These figures demonstrate the augmented efficiency of the American railway as a machine for the production of transportation or, expressed differently, that, as compared with its typical predecessor of 1880, 1890 or 1900 it is, today, of so much greater utility that it now enables the labor of one hundred men to accomplish what required the labor of 108 men in 1900, of 148 men in 1890, and of 194 in 1880. It ought further to be noted that these figures have no necessary relation to the quality of the labor employed or the labor services rendered, and afford no answer to any inquiry concerning either. If it may be assumed that such quality has neither deteriorated nor improved they plainly measure the results of improved machinery and organization. That assumption, is. perhaps, sufficiently warranted to be accepted at least for the purpose of the discussion that follows. Whoever objects to it can apply such modification as he may deem to be reasonable. It is customary to measure capital in terms of money, the standard of value, and hence the customary expression for a But particular quantity of capital is its equivalent in money. value is the current ratio at which commodities exchange for one another, and as these ratios fluctuate widely from time to time, comparisons based upon a long period may become highly deceptive on account of wholly unrelated and extraneous conditions. This is the case, if an attempt is made to compare the capital invested in American railways in 1880 with the amount now invested. The variable value standard has been so modified during this period, by economic changes, many of which were external to the matters under investigation that, if the attempt is made to relate capital thus measured to the amount of work done at the end of successive decades the resulting averages afford no suggestion of the increased utility of the unit of capital, this increase being concealed by (1) increased efficiency of railway organization, and (2) decreased cost of the separate units of the constructed plant and equipment, i. e., rails. locomotives, etc., making up the railway machine. That is to say, as compared with 1880, the railway tool of 1910 is more efficiently manipulated, and, reduced to comparable standards, costs less. The following table, comparing net railway capitalization with the work done in the transportation of freight shows what has happened. Year. 1880. 1890. 1900. 1910. 141,596,551,161 255,016,910,451 Net capitalization. 9,547,984,611 14,338,575,940 Average per ton of 15.71 10.54 6.74 5.62 The foregoing shows that for each ton of freight carried one mile in 1880 there was a capital investment, measured by the net capitalization, which is at least sufficiently accurate for the purposes of this paper, of 15.71 cents, and that by 1910 this had been reduced 64.23 per cent. to 5.62 cents. These ratios prove that the productivity of capital, according to the ordinary standard of measurement, as well as the productivity of labor has been multiplied, and suggest that during most, if not all, of the period the railway industry must have been in that state of conformity to the economic law of increasing returns during which additional "doses" of capital or labor or both, applied under capable management, give better results, in the form of a higher average ratio of volume of output to capital or labor employed, than could be obtained before. It is well understood, however, that this state cannot remain the permanent condition of any industry, and that at some stage or other of its development it must pass into the state of constant returns which will soon be followed, if continued expansion takes place, by the final state of diminishing returns. In this state it may still be possible to cheapen the average cost of output by substituting lower capital expenses for higher labor costs, but it will no longer be practicable to augment both investment and labor and to show a resultingly higher factor of efficiency for both. In the April, 1912, issue of Moody's Magazine there appears an article written by W. Martin Swift and entitled, “Railroad Operating Expenses," which is very suggestive along these lines. Mr. Swift's article is, in effect and so far as its data and con clusions are accurate and sound, a marshalling of evidence which tends to prove, first, that the railway industry of the United States, as an whole, is now subject to the law of diminishing returns; second, that the decline in the value of the separable units of its necessary physical property stopped some time ago, and, third, that of late it has been forced to purchase increased productivity of labor by greater investments of capital of relatively lower productivity than the average formerly attained by the smaller former investments. It must be noted that so long as the loss in average productivity of capital is less than the gain from the increased productivity of labor this And it must process has complete economic justification. further be noted that for the purpose of determining whether such justification exists all values of capital goods and all rates of wages must be reduced to comparable terms as the real question is whether day by day, under current operating conditions, the additional investment is justified. The computation is, perhaps, easier of comprehension than a statement of the principle. It might run like this: an additional capital investment costing, for interest and maintenance of property, $1,000,000 annually ought to be made if by making it total operating expenses, which are principally labor costs, at current rates of wages, can be reduced much more than $1,000,000 per annum. Mr. Swift uses the term of somewhat sinister suggestion, "over-capitalization," but any reader of his article perceives at once that he does not use it in the harsh and ordinary sense, but only as referring to the condition already described, in which more capital is necessary to produce an added volume of ransportation than would be required if the average productivity of capital of the past could be maintained. This article has already shown that this productivity was maintained and greatly increased, from 1880 to 1910, but the figures used by Mr. Swift indicate that the process had been reversed before the end of the period, and disregarding his statistics, which it has been found impossible in some cases to substantiate, this one of his conclusions is apparently sound. Referring to the last foregoing table it will be noted that beginning with a capital investment of 15.71 cents per annual tonmile in 1880 there was a rapid decline to 10.54 cents in 1890 and 6.74 cents in 1900 and that the slight decrease of the last decade carried the average of 1910 to 5.62 cents. Thus there was a decrease of 5.17 cents, or 32.91 per cent., in the first decade, one of 3.80 cents, or 36.05 per cent. in the second decade and one of 1.12 cents, or 16.62 per cent., in the third and last decade. A closer examination of the official data for the last decade shows that from 1901 to 1910 the productivity of railway labor has increased with substantial regularity but that the productivity of capital, considered alone and irrespective of its ability to advance the productivity of labor, has fluctuated and that, allowing for changes that are plainly due to the peculiar traffic conditions of different years, it has decreased during the last half of the period. These facts are made apparent by a close study of the following table, the basic data in which are from successive reports of the Interstate Commerce Commission, although a clearer presentation can be made, and will be offered hereinafter, by using two-year averages and thus avoiding some of the unrelated variations that result from using annual figures: Number of tons of freight carried one mile. Net capitalization. Number 131,974 10,281,598,305 Average per ton of freight carried one mile, in cents. 6.45 6.31 5.94 6.14 5.99 Average Not given by the Interstate Commerce Commission-obtained by assuming that the increase from 1906 to 1907 was one-half of the increase shown by the Commission for the years 1906 to 1908. |